Hierarchical Model Predictive Control for Energy-Aware Scheduling of Digital Twin-Based Batch Manufacturing Systems

Driven by growing concerns over global energy consumption, improving energy efficiency in the manufacturing sector is increasingly vital, especially for energy-intensive batch processes. Real-time pricing (RTP) of electricity, a dynamic demand-side management strategy, offers opportunities to reduce energy costs by scheduling production during low-price periods. However, integrating RTP into batch manufacturing scheduling introduces challenges in balancing energy savings with timely customer contract fulfillment, requiring multi-timescale coordination. This paper presents a hierarchical Model Predictive Control (MPC) framework integrated with a System-Level Energy-Efficiency Digital Twin (SLEE-DT) for energy-aware batch manufacturing scheduling. The SLEE-DT provides a unified system representation, capturing the dynamic interactions between production and inventory stages. The hierarchical MPC consists of two levels: an upper-level offline optimization that determines long-term inventory and production strategies, and a lower-level runtime controller that performs dynamic scheduling in response to system states and RTP signals. A case study of a battery production line demonstrates that the proposed framework reduces energy expenditures while maintaining reliable fulfillment of customer contracts, highlighting its potential for scalable, cost-efficient, and sustainable manufacturing operations.